Sperm Banks and Gays:

The Washington Post ran this story, written by its health reporter, on May 6:

A new Food and Drug Administration regulation scheduled to go into effect this month calls on sperm banks to reject anonymous donations from men who have been sexually active with other men within five years — an effort to screen out potential carriers of the AIDS virus.

With a May 25 implementation day approaching, the long-debated regulation is being attacked as unscientific and bigoted by some gay groups that say the new rule stigmatizes gay men.

"This rule is based on bad science because the AIDS epidemic is an increasingly heterosexual epidemic," said Kevin Cathcart, executive director of Lambda Legal Defense and Education Fund, which has asked the FDA to reconsider.

"We also know that sperm banks can test for HIV very well now and so can screen out any infected donor," he said. "With that in place, why are gay men being rejected?" . . .

The story goes on — I don't want to quote the whole thing, because of copyright reasons (there's a possible argument that quoting the entire article would be fair use, because I'm criticizing it, but I'd rather err on the side of caution). But nothing in the rest of the article says anything about what the FDA's response to this criticism might be, or about what (if anything) seems to be the conventional wisdom from objective and reputable sources on who's right here.

The closest the story comes to that is by saying "When the rule was adopted, acting FDA Commissioner Lester M. Crawford said it was developed with broad input and 'in all cases, we carefully considered the comments we received in the proposed rule and made changes in the final rule when the science supported the change.'" But what exactly does the science say?

It is pretty obvious — though perhaps some readers might miss it — that "This rule is based on bad science because the AIDS epidemic is an increasingly heterosexual epidemic" is an unsound argument. The risk that any U.S. male homosexual is infected with HIV are over 10 times (likely considerably over 10 times) the risk that any U.S. male heterosexual is infected with HIV. (I'm extrapolating from the year 2000 new AIDS infection numbers listed here, with a generous margin of error; I realize that this isn't the same as current HIV numbers, but it's close enough to get a confirmation of conventional wisdom, which is correct here. If anyone has more precise data, please let me know.) Given this, screening out those who are disproportionately likely to be at risk -- of course, as well as trying to use other methods to screen out donors who don't fit within the demographically most-at-risk categories, but who may have HIV nonetheless -- may well be sensible. If there's a flaw with the FDA policy, it is not that AIDS is an increasingly heterosexual epidemic (if it is that) even though in the U.S. it is still a highly disproportionately male homosexual epidemic.

But what about the argument that "We also know that sperm banks can test for HIV very well now and so can screen out any infected donor" — "[w]ith that in place, why are gay men being rejected?"? Are HIV tests certain enough that the sperm banks can efficiently and completely reliable test sperm, for instance, by freezing the sperm, testing the donor six months after the donation — six month seemingly being the period between infection and [nearly?] certain detectability — and using the sperm only if the donor comes up clean? Or is there some significant possibility of error even then?

In either case, wouldn't that information be useful in an article by a health reporter, more useful than just giving an advocacy group's take on the matter (plus a general assurance of safety from the operator of a sperm bank that accepts sperm from gays), with no sense of what impartial scientists actually say, and no discussion of any possible counterarguments?

I should stress that I genuinely have no views on what the science is. For obvious reasons, I'd love it if sperm banks could 100% reliably test sperm for HIV. It may well be that Mr. Cathcart's arguments are completely right.

But it seems to me that readers would like to have more than his arguments and those of a sperm bank operator that obviously has a stake in the matter. There should be some science on this out there somewhere, I take it, and some scientists who can speak to it, no? If there's a conventional wisdom on the subject that comes from highly reliable scientific sources, let's hear it. If the conventional wisdom is that we don't know what the right answer is, let's hear that. In any event, let's at least hear the specific arguments from the other side.

I am not trying to make any claims here about media bias. My point is simply that the story just isn't very useful to a typical casual reader who's trying to get an objective sense of the matter. Not a grand journalistic felony, I realize -- but it's the sort of journalistic misdemeanor (or even infraction, if you prefer) that, when repeated as often as this one seems to be, makes for unhelpful newspapers and ill-informed readers.

As always, please let me know if I've missed something important here.

Sperm Donations and Homosexuals:

I should stress again that my post this morning about the Washington Post article was not an attempt to explain who was right and who was wrong -- rather, it was a criticism of the article's failure to adequately explain both sides of the debate, and of its failure to summarize what the conventional scientific wisdom is on the subject.

For those who are interested in the substantive question, though, you can find what seem to be the controversial guidelines here (please let me know if I found the wrong document; I wish the Post online article had had a link to the right one). The guidelines suggest the elimination of a wide range of people who seem at risk for HIV (pp. 16-21), including male homosexuals, intravenous drug users, hemophiliacs, men who have patronized prostitutes (of any gender), and 16 other categories of people.

Interestingly, the guidelines do not include people who have had promiscuous heterosexual sex; I don't know whether the FDA didn't see this as a high enough risk factor for HIV transmission, or whether it had some other reason for this. Perhaps this is a reason to criticize the guidelines -- but if it is, then again it would have been nice if the Post article had discussed the issue.

You can also find the FDA's responses to comments -- again, I hope these are the right ones -- here. Here's what they say about the exclusion of homosexual donors:

Some comments disagreed with considering homosexual men to be "high risk donors" and disputed the scientific basis for excluding these men as donors. Many comments cited the efficacy of the blood test for HIV, with retesting after a 6-month quarantine, although one comment noted that HIV antibody testing is imperfect. Many comments disputed the public health benefits of the rule, although some applauded the agency for trying to craft safeguards to protect the public. . . .

In response to the comments suggesting that FDA should allow establishments to rely on HIV test results alone, or on quarantine and retesting, without screening for risk factors, FDA rejects that approach at this time. Although it is reasonable to expect that more sensitive nucleic acid amplification testing (NAT) will be available soon for reproductive tissue donors, even that testing may fail to detect early stage HIV and other infections, particularly because the level of viremia may be extremely low in the early stages of infection (Refs. 1, 2, and 3). Moreover, even the best test may fail to provide an accurate test result due to human error in running the test or in linking the test result to the correct donor. Accordingly, FDA believes that, based on the current state of testing and current knowledge about disease transmission, it is necessary to screen for risk factors as well as to test for diseases such as HIV. . . .

After the consultation, it was concluded that there is no new data that would warrant revising the 1994 guidelines. CDC and others also concluded that current data are not sufficient to allow the identification of lower-risk subsets of currently excluded population groups, and thus, to refine the exclusionary criteria. At the consultation, representatives of CDC encouraged the development of new data.

On December 14, 2001, we asked the Center for Biologics Evaluation and Research's (CBER) BPAC, whether there are existing data that identify subsets of men who have had sex with other men in which the incidence and prevalence rates for HIV, HBV, and HCV of the subsets are similar to the population at large. By a 10 to 0 vote, the committee advised that these data do not exist.

We have reviewed relevant legal authorities and disagree that these regulations discriminate or improperly abridge donor or recipient rights. We further note that, since FDA has tailored the rule's requirements to take into account an existing relationship between a donor and recipient (for example, FDA has not required quarantine and retesting for directed reproductive donors, permits the use of reproductive tissue from ineligible directed reproductive donors, and requires no testing for sexually intimate partners), the comments' remaining objections relate almost exclusively to anonymous donations of reproductive tissue. We will continue to examine the data on risk factors and, as new data are developed that justify changes to our guidance, we will make those changes in accordance with good guidance practice.

For more details, see the document. Again, I can't speak to whether the FDA got this right -- but wouldn't it have been good for the Washington Post article to explain a bit about the FDA's substantive defense of the guidelines (or, even if the FDA hadn't defended them, what those who defend the guidelines say)?

Current Relative Rates of HIV Infection:

A reader kindly passed along to me what seems to be the most recent, albeit geographically limited, data on the subject, from Texas Department of Health. The data seems to be 2003 data, based on then-living people in Texas with known HIV infections.

If you set aside IV drug users and hemophiliacs (I oversimplify slightly here) — groups that the FDA urges be excluded as donors, regardless of their sexual practices — and the unreported risk people, we have 7239 males that were apparently exposed through homosexual sex, and 920 through heterosexual sex. If we estimate that 4% of the male population is homosexual (the numbers that, to my knowledge, are most reliable), this means that the average male homosexual in Texas is nearly 200 times more likely to have HIV than the average male heterosexual; the rate for homosexuals is 7239/(.04*10,000,000) = 1.8%, while the rate for heterosexual is 920/(.96*10,000,000) = .0096%. (The male population of Texas is 10,000,000.) This is something of an oversimplification, and it's based on a limited sample. Still, I suspect that this is a decent back of the envelope calculation; if it's off by even a factor of 5, that's still at least a 40-fold higher risk.

HIV is a tragedy; I hope it gets cured as quickly as possible. But it's a mistake, I think, to deny that it's a tragedy that afflicts male homosexuals at not just a higher rate than male heterosexuals, but at a vastly higher rate.

Incidentally, some people asked why the FDA doesn't similarly exclude sperm donated by blacks, because they have a higher HIV infection rate. Analogies to race are sometimes helpful, but not always; but in any event, according to the Texas data, there were 5738 white males in Texas who had HIV, and 4820 African American males (I'm not sure whether they really mean African American in the sense of Americans of African descent, or whether they also include non-American blacks, but I suspect that for our purposes that doesn't matter much). Texas is about 11.5% black and 71% white, so the rate of HIV infections for black males in Texas is roughly 5 times greater than the rate for white males.

This omits all sorts of important factors, including the categorization of Hispanics and also the need to focus, once IV users and homosexuals are eliminated from the donor pool, on the relative rate of infection by race of non-IV-using heterosexuals, which may be different than the overall race-based infection rate. But even if we focus on the first-cut approximation, we find that being black, while a risk factor for HIV, is apparently a 30 times weaker risk factor than being homosexual.

A Bit More on HIV and Insemination:

A reader points me to this article,which reports:

Currently, 15 women are known to have been infected with HIV via artificial insemination using sperm from anonymous donors . . . [including] six in the United States. All but one of these cases of insemination-related infection occurred before the availability of HIV antibody testing.

The article (which is the first item shown by a google search for "hiv insemination") also reported that "approximately 75,000 women are artificially inseminated annually in the United States." If this is so -- and of course we'd want to know whether others disagree with these statistics -- then perhaps HIV screening is enough, and there is no need to disqualify homosexual donors. It's certainly possible; in fact, it would be excellent.

But again, the Washington Post article said nothing about this. This is interesting, useful information that actually helps readers think through the matter, especially if coupled with some sense of what other scientists think about this. But instead of getting this, the Post readers got something that looks more like a press release.